Working Paper: NBER ID: w27655
Authors: Shnke M. Bartram; Mark Grinblatt; Yoshio Nozawa
Abstract: Methodological insights generating comprehensive transaction-based bond datasets reveal that corporate bonds’ book-to-market ratios predict returns from prices transacted days after signal observation. Senior bonds (even investment-grade) with the 20% highest ratios outperform the 20% lowest by 3%–4% annually after non-parametrically controlling for numerous liquidity, default, microstructure, and priced-risk attributes: yield-to-maturity, structural model equity hedges, bid-ask-spread, duration/maturity, credit spread/rating, past returns, coupon, size, age, and industry. Spreads for all-bond samples are larger. An efficient bond market would not exhibit the observed decay in the ratio’s predictive efficacy with implementation delays, smaller yield-to-maturity spreads, or similar-sized spreads across differing liquidity/de-fault bond-types.
Keywords: Book-to-market; Corporate bonds; Mispricing; Returns
JEL Codes: G1; G11; G12; G14
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| book-to-market (BBM) ratios (G32) | corporate bond returns (G12) |
| book-to-market (BBM) ratios (G32) | predictive power of corporate bond returns (G12) |
| book-to-market (BBM) ratios (after controlling for liquidity, default risk, microstructure) (G32) | corporate bond returns (G12) |
| book-to-market (BBM) ratios (G32) | yield-to-maturity (E43) |
| book-to-market (BBM) ratios (G32) | risk-adjusted returns (G12) |
| BBM signal efficacy decays over time (L96) | mispricing accounts for BBM anomaly (G19) |
| book-to-market (BBM) ratios (G32) | robustness of predictive efficacy (C52) |